sign in sign up
<<
Home , 58 BC, Aedui, Allobroges, Arverni, Bellovaci, Bibracte

AEDUAN GAULISH ECONOMY AND SOCIETY ON THE EVE OF THE C0NQUEST:THE DIALECTICS OF LANDSCAPE AND POWER

Carole L. Crurnley

SPACE, TIME, AND SCALE A basic archaeological assumption is that temporal change is reflected in spatial change. Whether stylistic (as in the decoraticln of a bowl), or statistical and physical (measures of the spatial relationships in a site's accumulated deposits), archaeological evidence invariably has a spatial component. From spatial evidence temporal continuity and change are inferred. If archaeologists wish to examine the mechanisms of cultural change, rather than simply providing descriptions nf spatial and temporal variation in material culture. they must address the question of how cultural change is encoded in and inferred from spatial discontinuity. Fundamental to this efiort are analyses conducted simultaneously at muitipie scales (for an expanded discussion see Marquardt and Crumley 1987). Scale refers to the "grain" of the unit of analysis relative to the matrix as a whole; effective scale (Crumley 1979:164-65) is any scale at which pattem may be recognized and meaning inferred. For many years archaeology was dominared by research at ihe scale of the site, paralleling a focus in anthropology on community. In the second haif of the twentieth century, an emphasis on the individual and on context rendered legible severa1 more scales, ranging from regional settlement patterns to sub-site activity areas (Willey 1953; Binford and Binford 1968: Hil1 and Gunn 1977). Subsequent developrrients link concepts and techr~iques that allow the recovery of cartain cognitive aspects oí' vanished societies, such as the role of history and oi cultural preierence in adaptive strategies (Crumley 1993; McGovern 1994) or the recovery of aspects of landscape symbolism in vanished belief systems (Ashmore and Knapp 1996). In addition, increasingly sophisticated techniques for extracting more traditional economic, social, political, and environmental information have become available. Particularly fruitiul for the analysis of vanished polities has been long term regional-scale analysis. With the passage of time a polity, or its administrative successor(s) in a region. re-ranks the importance of that region's resources. These tluctuating boundaries. priorities, and perceptions necessitate a dynamic definition of region in which, like any artifact, its forrn and content is transformed over time. Even regions that at one scale are ostensibly homogeneous nonetheless have distinctive cultural and physical features, both within the area and overlapping into other contiguous areas; manifest at greater or lesser scales, these features harbor contradictions with the potential to negate any claim of uniformity. Thus the analysis oi regions and their political, social, and environmental boundaries and other divisions must be undertaken at severa1 temporal and spatial scales. The concept of lanrlscape. the spatial manifestation of the relations between human groups and their environments, is particularly useful to archaeologists. Landscapes offer a framework within which contemporaneous sites and olher zones oC activity can be irrtegrated with features of the physical erivirorimerrt (Crumley 1976:7; Marquardt and Crumley 19875; Crumley and Marquardt 1990). A landscape signature (ibid. 1987:7) is the material record, manifest in both physical and sociohistorical structures, of the human activities that characterize a particular time and place. From its signature, priorities, choices, and the interna1 dynamics and external relationships of a politp's administrative and political system rnay be inierred. Thus the concepts of region and landscape. in their transformation across time and space, and in their multiple meanings to those within and beyond fluctuating borders and divisions, are powerful tools for the study of pre- and prcto-historic polities. These integrated regional- and continental-scale data oiier new insights when applied to the topographical realities of the European countryside, and many factors in the decision-making process are clarified. The interpretative utility of this body of information, especially as it regards Gaulish polities, is

Cuadernos de Arqueologia Mediterrhnea. Tomo 3. 1997: 31-46 augmented by excellent research on [he patronclient relation (e,g., Eisenstadt and Roniger 1980; Kaufman 1974; Michie 1981; Powell 1970; Schmidt et al. 1977; Silverman 1965; Tarrow 1966; Weingrod 1968; Wolf 1966) and on party-based patronage, both in and at . Below I identify decision-making priorities as they pertain to the collective conduct of the various Gaulish polities and Rome and to the individual conduct of and Romans. Focus is on the Celtic Aeduii, a Gaulish polity located in what is now (Francel, and which I have argued is by any definitiol1 a siate (Crumley 1974, 1995a). Using textual evidence to begin [he analysis, I identify geo- and sociopolitical factors of major importance (priorities) during the period in question, and explore the ways in which the particularities 0f topography affected the ranking of those factors uppermost- - in the minds of the Celtic and Romans in the la& few centuries B.C. It can be argued that several important features of the sociopolitical organization of certain preconquest Gaulish polities and Republican Rome were essentially similar. The following discussion compares these twO structured systems in terms of shared, historically specific dynamic tendencies in political economy. Because topography (through definition of boundarles, and defense) records decisions in predominantly political and economic terms, the scale at which topography affects human affairs serves as an accurate indicator of systemic change (dynamic tendencies). The resultant anthropologicai analysis supports generalizing comments about the structural development of a ubiquitous element in European society, the patron-client relation.

GEOPOLITICAL AND ECONOMIC RELATIONS IN GAUL ON THE EVE OF THE CONQUEST By the time of Herodotus (ca. 484-424 B.C.) the Greeks recognized the Celts as a major people living west and north of the Mediterranean and beyond the . By the first century B.C., the boundaries of Celtic Gaul were well defined (conforming almost exactly to those of modern ), and (according to classical authors) the inhabitants distinguished themselves both from the Germans east of the and the south of the (cf. Wells 1997). As Caesar reports in De Bello Gullico, they also distinguished themselves internally as named ethnic polities. While the of Gaul at the time of the Roman conquest is of enduring interest (Blanchet 1931; Desjardins 1885; Holmes 1911; Jullian 1908; Longnon 1885; and many others), two scales of inquiry have predominat&: local (the battle, encampment, or line of march): and all-inclusive, the latter treating the many Gaulish polities as if their altercalions were merely a prirrtitive prelude to the Roman Pax. Only rarely (e.g., Büchsenschiitz 1984; Crumley and Marquardt 1987; Goudineau and Peyre 1993) has a single polity--its territory, economy, and relations with other polities and with Rome--been the subject of investigation. De Bel10 Gallico (BC), 's account of his six-year campaign (58-52 B.C.) in northwest Europe, is, however, full of information about named Gaulish polities and interpolity enmities and alliances. Caesar, correctly or not, also distinguishes among Roman, Celtic, and German machinations. While Caesar and other classical writers always pose interpretive difficulties (Christ 1995; Crumley 1974; Dobesch 1989; Dunham 1995; Rambaud 1953; Timpe 1986, 1989; Wells 1997), they also offer priceless information. Geologically, France is the portion of the European continent most marked by successive periods of mountain-building. Bounded on the east by the Alps and on the south by the somewhat older Pyrenees, the interior exhibits an even older orogeny in the we~tcentrairegion (the Massif Central) that dominates the geology of one-sixth the total area of France. AS a result, the topography of France presents a diverse program of deep gorges, rugged pinnaeles, windswept plains and coastlines, and gently rolling hills and valleys--frequently within sight of one another. IB abundant rivers drain the heights and flow into the Mediterranean, the Atlantic, the , and the North Sea. Its climate is predominantly temperate, although oceanic. continental, and Mediterranean weather patterns dominate in certain seasons or certain years (Crumley 1987a. 1993, 1994). One finds juxtaposed every type of soil, from windblown Pleistocene loess to granite and limestone-based soils. The major natural resources of Gaul were the direct result of this geologically and topographically heterogeneous landscape. Mountain-building activity insures the presence of a variety of metals (gold, silver, iron, and tin, among many others). The combination of fortuitous climate and soils offers a nearly infinite variety of agricultura1 milieux. Finally, and perhaps most important, Gaul's geographic situation in western Europe assured the importance of any group or groups that could control its pons, its rivers, and its At the time of Herodotus, Celtic peopies spanned the length of every major western European trade route, namely the coastal route between and . the Aquitanian Gap passage from the French L~~~~~to the Atlantic via the River, and the corridor dong the Rhdne and Sabne rivers. The third of these was particularly important: it connected with the River by a short, easy overland route and thence to the Atlantic; it marked the end of the major east-west pass across the Alps (at ); and it gave onto the Belfort Gap and and rivers, and thence to the entire Rhine basin. Moreover, the source of the , not far from the upper Sabne and reached by an easy upland crossing, gave access to the north coast and Britain. In short, the whole of western Europe was laid open to commerce as a result of the accommodating topography of Gaul. There was an intimate relationship between the boundaries of Gaulish polities and the means of livelihood of their inhabitants. A long-established European pattern of mixed agriculture and husbandry (Hubert 1934; Crumley 1994) would have been practiced in the more heterogeneous landscapes, and both intensive agriculture and exclusive cattle-, horse-, and sheep-raising in the more homogeneous river valleys and upland peneplains. Other aspects of the topography would have received particular attention, among them defensible high places, navigable stretches of rivers and river fords, springs (particularly the sources of major rivers), and mountain passes. A single polity controlled most such places for its exclusive religious, commercial, and administrative use. However, some strategic spots were situated at the boundary of ofie polity with another. The Loire, for instance, marked the boundary between the Aedui and the Bituriges Cubi; the Sabne, the boundary between the Aedui and the . In the first instance, the Bituriges were clients of the Aedui, and their compliance was assured by hostages and mutual interests. But the boundary between the Aedui and the Sequani was disputed, its history marked by armed conflict. More commonly, high ridges, swamps, forests, and other inhospitable terrain marked polity boundaries (see Jullian 1909:2:26ff.). In general it rnay be supposed that the central fortified town of a Gaulish polity was situated well within its boundaries, while its commercial towns were situated along major routes of passage and protected by armed garrisons. Bvth types of settlement were critical to the continued prosperity of the polity as a whole and, as such, were considerations foremost in the minds of the polity's leaders. The very nature of an overland trade route or navigable river is here at issue: such places serve simultaneously as edge and center. A boundary at a river not only divides two territories and serves to limit them both, but centralizes interaction between them and in turn links both territories to areas up- and down- stream. Cities at such boundaries serve the same function: they aggregate, integrate, and arbitrate varieties of custom and opinion. It rnay thus be said that boundaries and centers are similar in that they have simultaneously agglomerative and disagglomerative functions (Marquardt and Crumley 1987). Inevitably, the presence of a desirable resource at the boundary of two or niore polities would have led to disputes concerning utilization, exploitation, and taxation. Multiple claims of that nature may be adjudicated in four ways. First, the polities involved rnay mutually exploit the resource through treaty. Second, a powerful polity rnay simply usurp the right from a less powerful neighbor by force. Third, two equally powerful polities rnay dispute the holding for years, periodically exchanging control by force. Fourth, a more powerful polity rnay become the patron of a less powerful group, at which time control of the disputed resource rnay be left in the hands of the client, and taxes paid to the patron; or the client rnay be reimbursed as an employee for managing the resource or compensated in some other manner (perhaps protection from a mutual enemy). It has been noted quite frequently that the boundaries of Gaulish polities are difficult, if not impossible, to determine. I would argue that this is due to frequent renegotiatlon (friendly or not) of interpolity boundaries as each group's fortunes fluctuated. Such renegotiations rnay represent necessary adjustments subsequent to marriage alliances, friendly agreements between entrepreneurs of differing resources who decided to consolidate their interests and holdings, and changes in ownership through inheritance, bankruptcy, and the like. Such changes in control were tied to the variable fortunes of individuals and of groups. An excellent example of the effects of individual control of such a boundary resource is found in Caesar (BG 1.18). The Aeduan was said to have bought the right to collect Sabne River tolls and all other Aeduan taxes at a low price, "for no one dared oppose his bid." , elder brother of Dumnorix and a trusted Roman ally, complained to Caesar (1.20) that he had lent certain of his resources and his "strength" (probably some of his clients, that is, and the like) so that Dumnorix could establish himself. Now Dumnorix insisted on challenging Diviciacus for control of the Aedui, the two brothers having found themselves allied with opposing pro-Gaulish and pro-Roman factions. That no one dared oppose Dumnorix' bid was perhaps partially due to his punitive powers; but perhaps also (and more importantly) his countrymen saw the advantage of a powerful, pro-Gaulish patron at the limits of their polity. As for polities that control parts of the same trade route, it is clearly to their common advantage. along the route's length, to maintain (1) a moderate level of taxation so as not to send commerce elsewhere, and (2) reasonable access to would-be users. An example of such an agreement was that between the Segusiavi (upper Loire), the Ambarri (RhGne-Loire overland connection at MBcon), the Aedui (upper-middle Loire), the Bituriges Cubi (middle Loire), and the (overland routes in central Gaul). The Segusiavi, the Ambam, and the Bituriges Cubi were clients of the Aedui, the most powerful of the fou and most able to maintain free trade and general security for the entire region. In retum, the Ambarri, Segusiavi, Bituriges Cubi, and other Aeduan clients served in the Aeduan military and probably paid .~omeof the taxes collected by Dumnorix and mentioned in Caesar (BG 1.18). From the above, we may deduce the more obvious sources of wealth in Gaulish polities, most particularly those involved in long-distance trade with 0th policies or with Meditenanean entrepreneurs. Caesar (BG 1.1, 39; 3.1; 8.3,42, 55) notes numerous instances of the latter deep in the interior of Gaul. The sources of wealth were taxation of commerce, commerce itseif (import and/or export of both raw materials and finished products), and production for commercial enterprise.

Table 1 AEDUAN GEOECONOMY ONTHE EVE OF 1'HE CONQUEST

Sources uf wcalth Related resources Geograpliical location

Tnxolror~I?/ cur~rrrrcn.e control of passage at fords, passes, fords, etc. passes, brcak-~n-bulk points, Ieg~tirnacy

Irrl}wrl ~rrrtlc~kl)rlrl o/ cr~rrrrr~cr~c brokcrage rclations; stur- nudcs in transportation (intra- and intcrpol~tyand agclbreak-ili-bulk facilities nctwork international) Cor~rr~rercralproducltort stockbreeding horses access to excellent grazing lush valleys land cattle access to good graLlng land uplands, rolling hills sheep access to poor gra~ingland highlands pigs acccss to land for piglots corrals in forests rnelal-rnin~ngopcratlons acccss to nilncs Morvan liiglilands artisanal activit~es relations with artlsans andlor artisans' quarters, as at Ux- brokers as a patron, access ellodunum or to rnater~alsas practiclng artisan agr~culturslproduction access to good farrnland rlver valleys, roll~ngIlills, uplands Prulccliurl cavalrymen (arrned rncn and tlic~rllorses) defens~bleplaces hillfurts, other weapons brokerage relations (cornmer- cial, IcgaI, social, etc.)

Production might take a variety of forms: stock-breeding (cattle, sheep, and perhaps horses), control of metal-mining operations (BG 3.21; 7.22), control of artisans who turned raw materials into finished products (as with the patrons of Uxellodunum, BG 8.32, and Bibracte), and possibly larger agricultura1 ventures (either through control of large holdings or through ~00perativeand/or ~0ercivecontrol of individual farmers' smaller holdings). This last possible source of wealth is problematical, as it hinges on questions of inheritance and land tenure (sec below); but Caesar notes that in some supposedly friendly territories "the people" could nat be made to deliver grain to his armies, which implies that farmers were in control of their own produce (BG 1.16). Finally, the most obvious source of wealth, as a careful reading of De Bello Gallico indicates, was protection. Whether it was against another polity, Or against factions within a polity, patrons offered their clients physical protection from invaders and marauders as well as less tangible intercession on behalf of clients in disputes of administrative or personal nature. These sources of Gaulish wealth were protected by in specific resources, which in tum may be linked with some certainty to the physical geography of Gaul. Table 1 relates sources of wealth and resources in the territory controlled by the Aedui. The taxation of commerce (cf. BG 1.20) requires the armed control of mountain passes, river for&, and commercial break-in-bulk points, th~snecessitating a loyal and effective security force. In addition the right to tax must be legitimated (that is. sanctioned by. or a part of, the recognized government). Dumnorix1purchase through a bidding system, of the right to collect tolls on the SaGne River indicates his control both sufficien; funds for a large capital investment and a foiiowing of armed and mounted men. me import and export of commerce (intrapolity. interpolity, and internarional) requires investment in storage facilities at break-in-bulk points, as well as in boats, wagons, and draft animals. one would expect the consolidation of such resources at fords, passes, and break-in-buk nodes of the road network (Dowdle 1987). The agents (negotiants) of such enterprises would presumably (by working necessity) be skilled in cultural as well as economic brokerage. For most of Gaul, intrapolity commercial production focused on stock-breeding, metal-mining and -working operations, other artisanal activities such as enameling and ceramics, and agricultural production. Among the Aedui, the breeding of pigs, cattle, sheep, and horses is well documented (Charlesworth 1926; Davies 1935; Oaks 1987; MacKendrick 1987). Some metalextraction sites in Aeduan territory are known, as well as centers for smelting, plating, and enameling meta1 (Davies 1935; West 1935). Though some of the evidence dates to the first century after the Roman conquest, it is relatively safe to assume that on the eve of the conquest pottery, figurines, linen (among the Bituriges Cubi), wool, building materials such as timber and stone, and perhaps wine (Bituriges Cubi and Aedui) were both regularly exported and distributed locally. Recent research in historical linguistics finds many words for tools employed in contemporary French viticulture to be of Gaulish origin. There is evidence of preconquest agricultural production of wheat and other grains, vegetables, fruits, and nuts, as well as fodder and animal products (ham, cheese) in the territory under Aeduan control (Charlesworth 1926; Davies 1935). Centers sheltering artisans, and environments favoring husbandry and certain kinds of agricultural endeavor, are also known (MacKendrick 1987). Finally, the ability to provide for the protection of resources and capital and to insure personal safety necessitated cavalrymen (armed men and their horses) and defensible fortifications. Further, access to a complex network of social, economic, and political ties was mandatory and placed those whose business it was to defend the polity in an explicit relation with its entrepreneurs. In surn, geopolitical and econornic relations in Gaul in the FIrst century B.C. were firmly rooted in the physiognomy of the countryside as it related to the importance of commercial trade. The social and political relations were, to a large extent, the predictable outgrowth of resources being exploited at intrapolity, interpolity, and international scales.

SOCIOPOLITICAL RELATIONS IN GAUL ON THE EVE OF THE CONQUEST Just as the realities of terrain intruded on Celtic political relations, Celtic social organization stood in a dialectical relation to political and economic organization. With specific reference to preconquest Celtic society, the tem social structure denotes the means by which individual social status was granted, the relationship of status to social class, the nature of marital and familial relations, and the means by which strong supra-kin bonds were built through political liaisons. Elsewhere, in an extensive survey of the documentary evidence for Celtic social structure (Crumley 19741, I have synopsized what rnight be considered ethnographically correct about the continental Celts as judged by a representative group of scholars (Grenier, Holmes, Chadwick, Tierney, and many others) who have written in depth on the subject. From this exercise it became ciear to me that certain aspects of Celtic society might be spoken of quite confidently (social class, patron-client relations) whereas others (the alleged disappearance of kingship, rules of inheritance and of descent) might remain impenetrable if documentary evidence alone was employed. Our interdisciplinary project in southern Burgundy was begun in part to determine whether a close inspection of the physical (both material and spatial) evidence of a particular polity in Celtic society would shed light on these disputed areas of Celtic social structure. The condensed discussion that follows is a further step toward that end. A useful starting point is to distinguish between status (social and political) that is ascribed and status that is achieved. Ascribed status in Celtic society was obtained at birth and was dependent upon family standing, particular kin relations, gender, and social class. Families were strongly patriarchal (Grenier 1970 [1945]:191), and wealth was inherited bilaterally (BC 6.19). What is known about inheritance and descent is particularly problematic, for much information is gained through literary evidence from the British Isles and is of a significantly later date. It seems relatively safe to assume the existence of strong clan ties in the earlier continental polities, but distinctions in the inheritance status of women are apparent if earlier (BC 6.19) and later (cf. Binchy 1936) sources are contrasted. Women did not break ties with their own families on marriage, but continued to interact with them socially and politically, a circumstance that may be termed complementary filiation (Fortes 1945). Although the early Irish law tracts contain extensive information on the status of Celtic women in Christian Ireland, the information is of late date (sixth-seventh centuries A.D.) and difficult to use (but see Patterson 1995). One can, however, venture some conservative suggestions: for example, that Celtic women in the period before the conquest could hold and inherit property as individuals, but might be expected to with (and generally follow the wishes of) their male relatives (both affinal and agnatic) in mY decisions. *ence the importance of mamage alliances between families with common or c~mplementar~intc!rests. I have demonstrated (Cmmley 1974) that Celtic social class was undergoing a structural transformation at the time of the Roman conquest. Stimulation of the Celtic economy through long-distance trade (particularly with Mediterranean groups) had begun to blur class-based distinctions by offering individuals a vari et^ of WaYs to power, which at an earlier date only the aristocracy had enjoyed. The complex and enigmatic evidence for an earlier institution of kingship, perhaps ~ociatedwith the priestly class of dmids, invites close textual anal~sis of classical references to a Celtic aristocracy comprised, at the time of the conquest, of druides and equifes. Caesar (BG 6.14) notes that the , the more exalted of the two, were exempt from taxation and took no pan in warfare. It may be that all young aristocratic Gaulish men took some druidic training; other information suggests that an aristocracy existed apart from the more parochial activities of the druids. Certainly it would appear that the function of the aristocracy was predominantly administrative. Grenier (1970 [1945]:183, 216) suggests that the aristocracy was only a few generations old at the time of the conquest and had replaced a system of royalty that had fallen under the pressure of a class whose wealth increased through trade with the Mediterranean world; thus more individuals shared in control of the economy, and status differentiations were multiplied. Although the origins and functions of Celtic kngship arc hotl~ disputed, the above does remain a plausible explanation for the continued presence of "kings" among the British Celts and Celts in Aquitania and Belgica, all areas at the periphery of the most intense Gaulish commerce. To whatever we may attribute the disappearance of kingship, certain factions of the aristocracy would have found themselves strong enough to abandon higher allegiances. (4.4.3) reports that the Celts chose a magistrate annually, perhaps meaning for all Gaul; Caesar confirnls this (BC 6.13). Much evidence exists (BC 1.18, 43; 2.1; 5.6, 20, 25; 7.4, 31) that Caesar was aware of the effects of the shift away from kingship and used the ensuing instability of rule to his own (and Rome's) advantage whenever possible, without taking a position for or against the institution in Gaul (or, some would be tempted to say, in Rome as well). Although native social class remained the most basic asset in an individual's rise to power in Gaulish politics, by the first century B.C. social class and family ties alone were no longer sufficient to insure a powerful position in Gaulish society. Achieved status, specifically the personally controllable means by which one might increase status, became central to Gaulish sociopolitical organization. The most obvious use of both birth and personal initiative to achieve an enhanced status is through marriage with a member of another powerful family. The Aeduan Dumnorix (BG 1.9, 18) married the daughter of the Helvetian leader and likewise arranged the marriage of his mother to a powerful citizen of the Aeduan client group to the west, the Bituriges Cubi, thus strengthening ties with surrounding polities and rivaling his brother Diviciacus' power (1.20) despite the latter's support by Caesar and Rome. Personal qualities--generosity, or supreme daring and bravery in battle (1.18)--were also rewarded, for they emphasized values respected at all levels of Gaulish society. In addition, bonds of friendship (sometimes on the basis of swom oaths, 3.22, 8.33) and hospitality (1.47) cemented both intrapolity and interpolity relations. Most important, however, to gaining individual status and prestige was the complex system of Gaulish patronage. Although there has been much disagreement in the literature concerning what elements constitute the patronclient relation (discussed at length below), for the moment a deflnition in functionai terms will suffice: the patron offers economic assistance and protection from legal and illegal authority; in retum the client offers demonstrations of esteem, information on the machinations of others, and the promise of politica1 and rnilitary support. Apparently both dmids and served as patrons of the Gaulish lower class, but members of the lower class (Caesar refers to them by the at least loosely equivalent tem plebs) were nat necessarily always clients; certainly there existed wealthy commoners with followings of their own. The institution of clientage was the basis of Celtic political organization. Re earliest rnentiOn of clientship among the continental Celts is by (Historiae 2.17), who describes the advantages to a Gaulish noble of having a number of retainers and clients; Caesar (BG 6.15) also notes mutual advantages. of course the institution itself has a lengthy history among many Indo-European peoples: identifies a similar system among the Thracians, the Germans, and the of India (Mauss 1926). Grenier (1970 [1945]:220) remarks that one cannot exaggerate the importance of clien&hip in eaul, as it was the major source of power of such leaders as Orgetorix and Dumnorix. Other systems may have relied primarily on the solidarity of the gens (broad descent group); among the Celts the familia of a chieftain, relatives and retainers, was of greater importance. The word Polybius uses to describe is "brotherhood,' which tempts comparison with modem Sicilian social smcture. In the celtic example, the posited ascendancy of the patronclient relationship led to a more ~ophisticatedpolitical organization that cut across kin, class, and polity boundaries and that operated at every level of sociery. The power of particularly sirong patrons (such as Dumnorix) even rivaled that of the hiehest elected governmental officials (BG

In 58 B.C. , utilizing these contacts (BG 1.44), offered his clientship (particularly militar^ services) to Rome in retum for control of Gaul. But Caesar insisted that Gaul belonged to Rome and ~roceeded to prove his case to both the Gemans and the in subsequent campaigns. Amid all the intrapolity, interpoiity, and internatimal rivalries and accords that intensified as Gaul came under Roman conquest, only the Arvernian , by the power of his word and deed, argued for a single Gaulish state--"one policy for the whole of Gaul, whose unanimity not even the world could resist" 7.29.6, trans. Edwards 19 17). In light of the heterogeneity of Celtic politica1 factions, that Vercingetorix came as close as he did to achieving a united Gaul is an enduring monument to his vision, remaining unblemished by subsequent events. It is worth emphasizing here that although the Celts in Gaul had evidently recognized their geographic cultural unity, they had not been nationalized (as the peopies of the had, some two hundred years before) such that the affairs of constituent polities could be peacefully subordinated under a central authority (federation, republic, or empire) that would act internationally for the whole. We may see, then, among the Aedui theniselves and between the Aeduan and other Gaulish policies, as well as and Rome, the flexible power of the patronclient relation, crossing a variety of barriers: kin, class, polity, language, and culture. Its existence throughout western Europe and its elasticity at different scales demonstrated, we now turn to a short discussion of the Roman patron-client relation and then to a contextual analysis of the Roman and Gaulish forms of that relation in the first century B.C.

ROMAN ECONOMIC, POLITICAL, AND SOCIAL RELATIONS IN THE FIRST CENTURY B.C. Roman economic, political, and social relations in the first century B.C. are abundantly documented and afford an easy opportunity to compare and contrast the Roman patron-client relation with that of Gaul. Much rnay be learned from classical texts and inscriptions, and many eminent nineteenth- and twentieth-century historians (e.g., Mommsen 1968; Gelzer 1912; Syme 1939; Badian 1968; Taylor 1966) have addressed the subject at length. One must of course keep in mind the effect of the rich written tradition in Roman history, which gives the illusion of deepening the gap between Rome and Celtic society, whose profoundly oral traditions, now lost, are further obscured by the effects of subsequent Roman acculturation. In fact, closer comparative analysis discloses many similarities as well as some notable differences (Dunham 1995). Sources of individual status and prestige in Roman society may also be described in terms of ascription and achievement. Ascribed status for the Roman citizen was obtained at birth and was dependent on family standing, kinship, gender, and social class. Indeed the other factors might reasonably be subsumed under the category of class, for a hierarchical Roman class structure was well developed by the first century B.C., and noble families (patricians) were ranked on the basis of the prestige of their respective ancestors. Classes within classes existed among the nobility (Taylor 1966:26). Descent and inheritarlce were reckoned bilaterally, but in these great families the individual's social role was determined by gender. Noble men were expected to advance rapidly through a ranked succession of political offices; noble women were expected to aid their kinsmen (affinal and consanguinial) in that task through advantageous marriage and interstitial social and political influence. The knightly class (equites, men with property valued at mare than 400,000 sesterces), a third class that had evolved between the patricians and the plebs on the basis of wealth and (later) political interests, was likewise internally ranked. Thus ascribed social class depended first on birth; then status within the class was determined by prestige of the family and its wealth and influence. ClasS distinetions could be (and fairly often were) bridged by marriage and adoption. The closest bonds between noble houses were established in the same mariner (Taylor 1966). Adoptions were particuiarly important, as a male individual might thereby augment bis wealth and social standing without necessarily involving other family members (Goody 1969). After that there remained measures of p~relyachieved StatUS: personal political and, indirectly, economic success. Political success in public office was closely connected to bonds of friendship between families and to the Roman system of patronage. Nobles inherited both friends and enemies and were constantly engaged in making advantageous new contacts. They also inherited clients, and groups of friends and clients fomed fie basis of the Roman political party system by delivering popular support Theoretically, the client entered into the amngement voluntarily, trusting (in a relationship of fides) the patron to interpret the law to him and defend his interests, while the client's responsibility was to serve the patron in every way. The Greek historian Dionysus of Halicarnassus (2.9-1 1) provides a detailed discussion of Roman patronclient relations and notes that it waS unlawful as well as impious for patron and ,-lient to accuse one another in lawsuits or give votes or witness against each other and that whoever was convicted of activity was guilty of treason. He also stresses the political importance of the client in the noble's drive for political advancement in the cursus-honorum, the prescribed succession of political offices: ideally, quaestor at about thirty years of age, praetor nine years lates, then consul three years after that, after which proconsulships (provincial governorships) were regularly awarded. Between periods of service in these offices and others, a noble was expected to distinguish himself in military service somewhere in the provinces. It was by this means that a following of clients and friends at Rome was e~~hanuedby the allegiance of a loyal and well-tsained fighting force, individuals who themselves had been attracted by the hope of land and spoils. Thus a noble with a strong family background and widespread ties of friendship and patronage could advance rapidly in Roman society of the first century B.C. Julius Caesar is himself an excellent example of such an individual, and his campaign in Gaul served to enhance his personal standing as much as to press Roman economic interests in western Europe. Caesar supplies the classic case of how individual ambition and the goals of an expansionisr state coincide to produce a situation where both . While skillfully manipulating patron-client relations within Gaulish society (as with the Aedui and the ) and winning the ever-increasing loyalty of his army. Caesar insured the gracious receipt in Rorne of the news of his victories by securing for his pairia the two ma,jor trade routes in western Europe that had remained in Celtic control: the route Srom the Alps to the Rhiine and westward toward the Loire, and the north-south route that linked the Atlantic coast and the Rhine with the Rhcne. (See Figures A1-A6.) Caesar thus returned to Rome with more favors to bcstow and more individuals and factions in his debt. In fact his personal power had been a growing source of dismay to the Senate, who rightlp ioresaw that his political rivalry with (who had been similarly occupied in amassing power on a grand scale) might bring down the Republic. From Gaul, Caesar crossed the Rubicon into civil war. from which he emesged the first sole ruler of Rome in some 450 years. In 44 his assassins, whatcver their own private ambitions for power, nonetheless recognized that hierarchical patronage carried to its logical conclusion resulted in autocracy, essentially an extension of the kingship whose abolition had been an enduring source of Roman pride. Figures Al-A6: - Al. Control of the RhBne corridor, 122 B.C.: were subdued in 123, arc clients of Arverrli; Saluvii and Allobroges are allied. - A2. Control of SaBne, Seine, Loire, RhBneLoire corridor, 58 B.C.: and Sequani purchased German intervention in 71 to subdue Aedui; Aedui and Sequani disputed Sahne toll rights in 61, Sequani again usin:: German aid to OuSC Aedui; became Aeduan allies ca. 61; now, in 58, Sequani and mulually exchange hostages, Aedui unilaterally submit hostages to Sequani. - A3. "All were conspiring": control of the Rhine basin (Moselle, Meuse), the CalaisDover channel crossing, and the English Channel, 57 B.C.; among the Belgae the Bellovaci are known allies of the Aedui. - A4. Control of the English Channel and Atlantic coast toward Spain, 56 B.C.: Roman fleet campaigns against ; Crassus subdues Aquitania (Convenae here include several polities named in BG 3.27). - A5. Maintaining the CalaisDover crossing, controlling the mouth of the Rhine, 54- 53 B.C. - A6 "All Gaul" at Bibracte, 52 B.C.: Aedui begin as allies of Rome, then come under Arvernian, panGaulish alliance led by Vercingetorix; and Remi abstain, are fighting Germans to the east.

COMPARISON OF THE TWO STRUCTURED SYSTEMS As brielly described carlier, the rights, duties. privileges, and obligations that a patron-client relationship entailed wei.e quite similarly defined in Gaul and Rome: the patron offered economic assistance and protection from legal and illegal authority; the client offered demonstrations of esteem, information on the machinations of others, and the promise of political and military support. In both cases this relation forrned the basis for political parties and was fundamental to political organiration in general. Personal power, in both instances, was determined by the strength and extent of a loyal following (clients. friends, and retainers) and rhe connections an individual could clairrl i11 lligher ciscles. 30th Ro~nt:arid our Gaulish example, rhe Aedui, had a republican governmcntal organization with elected officials and laws governing succession to office. Both systems had been previously governed by kings (that is, dominance of government by a single family through inheritance and succession), the Romans in relatively remote times of which they retained a painful memory. the Gaulish polities in the somewhat niore recent past. Though both Roman and Gaulish patronclient relations were reflected in their respective legal codes and wcre the basis for political parties and personal power, they nonetheless exhibited some substantial differences, In Rome patronclient relations were highly formalized; among the Gaulish polities, more diverse. The above analysis of the Celtic system has yielded ample indications oí' the right of clients (in certain circumstances) to change patrons, and of the fierceness with which this freedom of choice was defended for every class. Although vicissitudes in this relationship were known in Roman society as well, they were considered exceptions rather than tbe rule (cf. 's support of Pompey, discussed in Taylor 1966:36). The lack of strict regulation of patron-client relations between Celtic polities is clearly indicated by the necessity for the transfer of hostages. Ties of friendship (equal status) between two polities were marked by the exchange of hostages (e&. BC 2.1, among thc Belpae); patron-client ties (unequal status) were marked by [he transfer to the patron polity of hostages from the client polity (e.g., BC 1.31. from the Aedui to the Sequani) (See Fipure A2). The core concep1 is one of bonding-bondage of an agseement between parties to insure confonnance. Another informal aspect of Gaulish patronclient relations was the much looser ranking of families and individuals, whose status wu subject to change and dispute. This apparent informality had no close analogue in Rome. where although family status was a subject of great interest, disputes concerning an individualts rnerit rarely included challenges to the illustriousness of his or her lineage, thanks to extensive genealogical records and scholarship. Perhaps it was a product of the lack of centralized conventions in Gaul. The elaborate dmidic oral traditions must likewise have included sophisticated genealogical lore; but inter-polity marriages, no1 to mention the decline of formal kingship, had over time cncouraged interrelationships of power So complex as to be defined almost situationally, both within and between polities. Both Celts and Romans might inherit relations of friendship and patronage. but far mare than their ,nore stable counterparts. the Celts were constantly obliged to renegotiate relations with friends and clients in ever-chaneing circumstances. L Although pauonclient relations were important at every level in both societies in the firs[ centun: B.C., the scale of patron-client relations (that is to say. the highest political and administrative organiLational level at which they regularly operated) was predominantly intra- and interpolity in Gaul, with incidences of (the druids' Council of all Gaul) and internarional (e.g.. "Friend oof Rome") accords. Ronia, whose illtemal politics were by this time essentially unified and stable, had extended its patronc1ie:lt system on the national and international levels as its provinces and allies accrued and its foreign policy became correspondingly more complex. Structurally speaking, Gaulish patronclient relations remained, as a result of topography and changing economic and social conditions, a predominantly heterarchical form. That is to say, patron-client relations sometimes took a hierarchical forn (a patron was in tum a client at a broader scale), buc at another scale relations among individuals might be structured heterarchically (on the basis of acknowledged equivalent, but not necessarily comparable, sources of power). For example, a patron of artisans (e.g., weapon-makers) might be the friend or the client or the patron of an individual who had complementary resources at his disposa1 (perhaps an efficient force of armed men, or an iron mine). Similarly, individual polities might likewise have complex links with other polities; at all scales, shifting needs in the areas of defense and economy, as well as the interference of other nations in the interna1 affairs of Gaul (e.g., BG 1.44, 5.27, 6.12) were sufficient to keep Gaulish patron-client relations from taking on a more formal, hierarchical structure. Both the Gaulish and the Roman state may be characterized in this period as resting (despite republican sentiments) on a basis of personal power through patronage. In Gaul, patronclient relations facilitated societal integration at a macro scale (as with alliances between polities). The personal nature of the patron-client relation also facilitated a reassertion of social, economic, and political relations at a more micro scale when the empire collapsed some 450 years after Caesar's conquests (Crumley and Marquardt 1987).

DISCUSSION I define elsewhere (Crumley 1979:144; 1987b; 1995b) two types of organizaiional structure pertinent to the following analysis: thc hierarchical, in which, on the basis of certain factors, some elements of the structure are subordinate to others and may be ranked, and the heterarchical, in which each element possesses the potential of being unranked (relative to other elements) or ranked in a number of different ways, depending on systemic requirements. We may draw from biology a simple example of the distinction: the Linnean classification system of the morphology of living organisms is organized hierarchically; but an ecological classification of species in an environment is heterarchically organized, such that various species may for a time be attributed equal importance as the system's components and parameters are studied, or one species may be designated more important than another as the focus of research changes. We may argue similarly regarding social structures: an individual's rank in a professional hierarchy, for example., may differ markedly from his or her relative status in a religious or community context. The individual's severa1 statuses in a variety of roles would thus represent various hierarchies, yet all might prove irrelevant in a court of law that upholds a code of uniform civil rights. The importance of this concept of heterarchy is that it reminds us that status in many societies is a shifting amalgam of ranks (Cancian 1965) whose individual and collective meanings (status) from time to time must be renegotiated. Periods of intense renegotiation predictably occur when the economy is stimulated and new opportunities for capital investment arise. "Wew money" Suels new investments, which further consolidate new status. Thus the econonly is doubly fed: both personal and corporate investments must be made wisely in order to maintain, consolidate, and enhance status. Astute decision-making ability (an acquircd quality, one assumes, distributed more or less randomly throughout the population) and the informed ranking of priorities (which assumes access to accurate informatim)--and not the ascribed characteristics of class, family, or gender-- become paths to power, and politization is complete. A key element is the patron-client relation: a workforce wishing to see its individual (or collective) well-being assured (clients), focuses infornlation in the hands of certain individuals (patrons), which facilitates the turning of resources into profits. In this context Caesar's argument that the Celtic lower classes were universally destitute and oppressed must be questioned not only on the basis of their freedom to change patrons and withhold compliance but on socioeconomic grounds as well. It may well be that there were two types of clients: those who served through clear perception of mutual interests (ambacti, soldurii) and others who were financially in a patron's debt (cf. BG 6.13-15). While yielding the point made in BC, I nonetheless find it difficult to envision very many circumstances in which mutual interests would not be ultimateiy financial. Frorn the above, as well as a single comment in BG 7.14, I argue that most Aeduans in the first century B.C. held property in private. Certain groups (e.g., artisans) may have held their capital in jewelry and stocks of their own rnerchandise more frequently than others whose wealth was derived frorn landholdings. Patriarchal descent and primogeniture would have kept landholdings in a descent group, and double inheritance would have assured women access to capital in the form of belongings and occasionally (as in the case of female only children) in land, although it is likely that men held and disposed of land more frequendy than women. Mamages between families providing access to both land and capital would have been propitious indeed.

CONCLUSIONS The factors influencing shifts in effective scale of patron-client relations during the past two thousand years, along with the shifting patterns of population agglomeration and dispersion, and of boundary maint~~ance and dissolution, are the subject of over two decades of investigation by our research group in southern Burgund~, once the heartland of the Celtic Aedui. An important piece of missing information in the construction of a model of first-century B.C. settlement and land use is the nature of Celtic rules of descent and inheritance practices, although careful work on the Irish texts may be able to remedy this situation (Patterson 1995). The present analysis of Aeduan geo- and sociopolitical organization in the period just before the Roman conquest has enabled us to deduce certain rules of descent and inheritance and to identify certain priorities, thus illuminating the decision-making process. Economic and social success would be measured in terms of increased access to resources. Indebtedness would mark the failure of a portion of the population to rank priorities effectively or to forn mutually beneficial liaisons, and would have been due to a variety of factors. Political, social, economic, and climatic instability would have required the borrowing of capital, although the influence of those factors would have been felt differentially. For exarnple, a highland mixed agriculturalist would have been less affected by a drought than a midlands grain or cattle farmer whose investments were in a few areas sensitive to climatic instability. To bring the best price, a metalworker with a specialty in enameling might have found it necessary to se11 his goods with others' en gros through a patron: another metalsmith might have followed an itinerant route from fam to farm, repairing metal weapons and tools. The social, political, and economic bargains struck with other, more powerful members of the society would have reflected each individual's situation (skills, circumstances of birth, etc.) and personal ranking of priorities within that existing framework. The foregoing analysis facilitates informed speculation about where the holdings of certain classes of Aeduan society (as well as those af specific individuals) might be found, (Crumley 1987c, note 2) relating social class to the landscape of the Aeduan polity. Too often we envision prehistoric and historic settlernent as rnonolithic--a single interlocking form, a hierarchical spatial representation to conform to our (also frequently erroneous) notions of a rigid social hierarchy. All about us today we have abundant evidence to the contrary: periods of rapid change invariably produce a variety of responses in all facets of society, not the least of which is the novel use of space (Crumley 1987c, 1993. 1994). The central importance of the patronclient relation in preconquest Gaul has been demonstrated. There is evidence pointing to its structural transformation from informal, heterarchical in the Iran Age, to informal, hierarchical under Roman influence in the Gallo-Roman period, to formal. hierarchical in the manorial context (Berry 1987). From Gaul in the first century B.C. to Merovingian France. the effective scale (the scale at which certain features forn a recognizably homogeneous unit) of patronclient relations fluctuates over time from local and regional to national and international and back to regional and local, in response to the fortunes of war and of longdistance trade. A final point must be made: in preconquest Gaul any status gained in the socioeconomic realm could be rendered irrelevant in the face of the power of the druids. Masters of information both secular and sacred (BG 6.20) and venerated by all, they served as social. political, and religious arbiters. Through their decisions, they focused [he energies and abilities of the Cel@ on service to [he mystic forces of [he universe, in retun, for *e imrnortality of the soul. It is unsurprising that the druids posed a threat SO profound that the Roman state moved to suppress their activities; the druids saw all too clearly that the power of the spring and the grove were to be supplanted by the lures of the marketplace and the scratch of the accountant's pen. APPENDIX: GAULISH DEFENSE OF TRADE ROUTES DURING THE CONQUEST The route to Spain and to the Atlantic through Aquitania had been secured for Rome in 121 B.C., with the establishment of the Provincia Narbonensis along the south coast of Gaul. The proposed emigration of the Helvetii in 58 B.C. posed a dual threat to that asset. They originally planned to take the route westward through the ; this Caesar firmly refused them, as too dangerous to interna1 stability (BG 1.6-7). The alternate, more northem itinerary through independent Gaulish territories promised widespread unrest that might threaten the security of the province from without (BG 1.10). This latter road was part of rhe great route that ran from the Alps to the Rhbne and westward. Indeed the subdued Ailobroges along the lower Rhbne had become restless, and the Aedui, Sequani, and Helvetii threatened to ally to close the upper Rhane (and hence the heart of Europe) to Roman trade (Figure A2). That threat was removed for the moment with Caesar's liquidation of the Helvetii, though a force sent by Caesar to protect the passage from the Rhone to the Alps was obliged to retreat after heavy tlghting with other mountain tribes (BC 3.1-6). In 57 (BC 2) the polities of the western Rhine basin attempted to maintain Celtic (more precisely, Belgic) control of the Rhine and its western tributaries (Figure A3), and in 56 (BG 3.7-19) the northern coastal polities attempted to maintain control of the English Channel and North Sea trade routes (Figure A4). Both groups of Gaulish alliances--the eastcentral and the northern and coastal--met with failure at the hands of the Roman armies, but British Celts offered more effective resistance in 56 and 55, because of Caesar's overextended supply lines and the difficulty of adapting Roman forces to local battle conditions (BC 4.20-35, 5.1-23). It was not unti1 52 (BG 7) that Gaul, under Vercingetorix. was able to make a united stand against widespread Roman control of European trade (Figure A6), and by then it was too late. Political boundaries shown in Figures Al-A6 are drawn from Lognon (1885; cf. Crumley 1974). In each figure, polities shaded in as alliances are mentioned specifically as such by Caesar in De Bello Callico, which is arranged by yearly campaigns. It can be assumed that adjacent and outlying polities were involved to a greater or lesser extent through patron-client obligations.

Carole L. Crumle?.. Departntent of Anthropology. Universiv nf North Carolina. Chapel Hill, NC 27599-31 15 USA

REFERENCES

Ashmore, W. and B. Knapp 1996 Sacred Landscapes: Constructed and Conceptualized. Symposium organized for the 1996 American Anthropological Association meetings, San Francisco. Badian, E. 1968 Foreign clientelae (264-70 B.C.). Oxford: Oxford University Pres. Berry, W. E. 1987 Southern Burgundy in Late Antiquity and the Middle Ages. Regional Dynan~ics:Burgundini? Landscapes in Historical Perspective: C. L. Crumley and W. H. Marquardt, eds.. pp. 447-607. San Diego: Academic Press. Binchy, D. A., ed. 1936 Studies in early Irish law. Dublin: Royal Irish Academy. Binford, L. R. and S. R. 1968 New Perspectives in Archaeology. Chicago: Aldine. Blanchet, A 1931 Carte archéologique gallo-rornaine. : Hachette. Büchsenschütz, O. 1984 Structures d'hahitats et forrifications de lAge du Fer en France septentrionale. MCmoires de la SociCtC PrChistorique Franqaise. Paris: SociCt.6 PrChistorique Franqaise. Caesar, Gaius Julius 1917 The Gallic War (De Bello Gallico). Translated by H. J. Edwards. . Cambridge: Harvard University Press; London: Heinemann. 1980 The Bottle for Gaul (De Beiio Gullico). Translated by A. and P. Wiseman with illustrations selected by B. Cunliffe. Boston: Godine. Cancian, F. 1965 Economics and prestige in a Maya community: the religious carga system in Zinacantan. Stanford: Stanford University Press. Chadwick, N. 1966 The Druids. Cardiff: University of Press. Charlesworth, M. P. 1926 Trade-routes and commerce ofthe Ro- Empire. 2nd ed., rev. London. Reprinted Chicago: Ares, 1974. Christ, K 1995 Caesar und die Geschichte. Hisrorische Inter~retationen.M. Weinmann-Walser, ed., PP. 9-22. Stuttgafl: Franz Steiner. Crumley, C. L. 1974 Celtic social structure: the generarion of archaeologically testable hypotheses fram literary evidence. Anthropological Papers, 54. Ann Arbor: Museum of Anthropology, University of Michigan. 1976 Toward a Locational Definition of State Systems of Settlement. Americun Anthropologist 78:59-73. 1979 Three locational models: an epistemological assessment for anthropology and archaeology. Mvances in Archaeological Method and Theory, (2: 141-173). M. B. Schiffer, ed. New York: Academic Press. 1987a Historical Ecology. Regional Dynamics: Burgundian Landscapes in Historical Perspective, C. L. Cmmley and W. H. Marquardt, eds., pp. 237-264. San Diego: Academic Press. 1987b A Dialectical Critique of Hierarchy. Power Relations and State Formation, T. C. Patterson and C. W. Gailey, eds., pp. 155-169. Washington DC: American Anthropological Association. 1987c Celtic Settlement before the Conquest: The Dialectics of Landscape and Power. Regional Dynamics: Burgiindiun í.undscupes in Historical Perspective, C. L. Crumley and W. H. Marquardt. eds., pp. 403-429. San Diego: Academic Press. 1993 Analyzing Historic Ecotonal Shifts. Ecological Applications 3(3):377-384. 1994 The Ecology of Conquest: Contrasting Agropastoral and Agricultura1 Societies' Adaptation to Climatic Change. Historical Ecology: Cultural Knowledge and Changing Landscapes, Carole L. Crumley, ed., pp. 183-201. Santa Fe: School of American Research. 1995a Building an Historical Ecology of Gaulish Polities. Celtic Chiefdorn, Celtic Stute, B. Arnold and D. B. Gibson, eds., pp. 26-33. Cambridge: Cambridge University Press. 195b Heterarchy and the Analysis of Complex Societies. Heterarchy and the Analysis uf Complex Societies, R M. Ehrenreich, C. L. Crumley, and J. E. Levy, eds. Archaeological Papers of the American Anthropological Association 6: 1-6. Arlington, VA: American Anthropological Association. Crumley, C. L. and W. H. Marquardt 1987 Regional Dynamics in Burgundy. Regional Dynamics: Burgundian Landscapes in Historical Perspective, C. L. Crumley and W. H. Marquardt, eds., pp. 609-623. San Diepo: Academic Press. 1990 Landscape: A Unifying Concept in Regional Analysis. Interprefing Spuce: GIS and Archaeology, K. M. Allen, S. W. Green, and E. B. W. Zubrow, eds. London: Taylor & Francis. Davies, O. 1935 Ronlnn mines in Europe. Oxford: Clarendon Press. Desjardins, E. 1885 Géographie de la Caule romaine. Vols. 2-3. Paris: Hachette. Dobesch, G. 1989 Caesar als Ethnograph. Wiener humanistische Blatter 3(1):18-5 1. Dowdle, J. E. 1987 Road Networks and Exchange Systems in the Aeduan , 300 B.C. - A.D. 300. ~~~i~~~lD~~~~~~~: Burgundian Land3-capes in Hisforical Perspterive, C. L. Crumley and W. H. Marquar&, eds., 265-294. San Diego: Academic Press. Dunham, S. B. 1995 Caesar's Perception of Galiic social structures. Celfic chiefdom, Celric state, B. Arnold and D, B. ~ib~~~, eds., pp. 110-1 15. Cambridge: Cambridge University Press. Eisenstadt. S: N.. and L. Roniger 1980 pationsiient relations as a model of structufing social exchange Comparative Studies in Sociev and History 22( 1): 42-77. Fones, M. 1945 The web of kinship arnong rhe Tallensi. Oxford: Oxford University Press. Gelzer, M. 1912 Die Nobilitat der riimischen Republik. and Berlin: Teubner. Goody, J. 1969 Adoption in cross-cultural perspective. Comparative Studies in Society and History 1 l(1): 55-78. Goudineau, C. and C. Peyre 1993 Bibracte et les Eduens: ci la décourverte d'un peuple gaulois. Collection Hauts Lieux dtHistoire. Paris: Editions ErranceICentre arch6010gique europCen du Mont-Beuvray. Grenier, A. 1970 Les Gaulois. Paperbound reissue of the 2nd (1945) ed., with new bibliography and preface by L. Harmand. Paris: Payot. Hill, J. N. and J. D. Gunn, eds. 1977 The Individual in Prehistory. New York: Academic Press. Holmes, T. R. 1911 Caesar's conguest of Gaul. Oxford: Oxford University Press. Reprinted, New York: Arco, 1971. Hubert, H. 1934 The Greatness and Decline of the Celts. London: Kegan Paul. Reprinted London: Constable, 1987. Jullian, C. 1908-26 Histoire de la Gaule. 8 vols. Paris: Hachette. Kaufman, R. R 1974 The patron-client concept and inacro-politics: prospects and problems. Comparative Studies in Society and History 16(3): 284-308. Lognon, A. H. 1885 Atlas historique de la France, depuis César jusquh nos jours. Paris: Hachette. MacKendrick, P. 1987 The Romans i11 Burgundy. Regional enamics: Burgundiun Landscapes in Histoatcal Perspective, C. L. Crumley and W. H. Marquardt, eds., pp. 431-445. San Diego: Academic Press. Marquardt, W. H. and C. L. Crumley 1987 Theoretical Issues in the Analysis of Spatial Patterning. Regional Dynamics: Burgundian Landscapes in Historical Perspective, C. L. Crumley and W. H. Marquardt, eds., pp. 1-18. San Diego: Academic Press. Mauss, M. 1926 Sur un texte de Posidonius. Revue Celtique 41: 497. McGovern, T. 1994 Management for Extinction in Norse Greenland. Historical Ecology: Cultural Knowledge and Changing Landscapes, pp. 127-154. Santa Fe: School of American Research. Michie, B. H. 1981 The transformation of agrarian patron-client relations: illustrations from India. American Ethnologist 8(1): 21-40. Mommsen, T. 1968 The provinces of the , vol. 5, Book 8: The European provinces. Translated by T. R. S. Broughton. Chicago: University of Chicago Press. Oaks, L. S. 1987 in the Aeduan Landscape: Transfunctional Deity under Changing Rule. Regional Dynamics: Burgundian Landscapes in Historical Perspective, C. L. Crumley and W. H. Marquardt, eds., pp. 295-333. San Diego: Academic Press. Patterson, N. T. 1995 Clans are not primordial: pre-Viking Irish society and the modelling of pre-Roman societies in northern Europe. Celtic chiefdom, Celtic state,B. Arnold and D. B. Gibson, eds., pp. 129-136. Cambridge: Cambridge University Press. Poly bius 1962 The (Historiae). Translated by E. S. Shuckburgh with an introduction by F. Walbank. 2 vols. Bloomington: Indiana University Press. Powell, J. D. 1970 Peasant society and clientist politics. American Political Science Review 64(2): 41 1-426. Rambaud, M. 1953 Lhrt de déformation historique dans les Commentaires de César. Annales de 1'UniversitC de , ser. 3, fasc. 23. Paris: Les Belles Lettres. 2nd rev. ed., Paris, 1966. Schmidt, S. W., et al. 1977 Friends, followers, and factions. Berkeley and Los Angeles: Unlversity of California Press. Silverman, S. 1965 Patronage and community: national relationships in central Italy. Ethnology 4(2): 172-190. Syme, R. 1939 The Roman revolution. Oxford: Oxford University Press. Tarrow, S. G. 1966 Peasant communism in southem Italy. New Haven and London: Yale University Press. Taylor, L. R 1966 Party politics in the age of Caesar. Berkeley and Los Angeles: University of California Press. Tierney, J.J. 1960 The celtic ethnography of Posidonius. Proceedings of the Royal Irish Academy 60, section C: 187-275. Timpe, D. 1986 Ethnologische BERgriffsbildung in der Antike. Germanenprobleme in heutiger Sicht, H. Beck, ed., PP. 22- 40. Berlin: Walter de Gruyter. 1989 Entdeckungsgeschichte. Reallexikon der gernzanischen Altertum~kunde7: 337-389. Berlin: Walter de Gruyter Weingrod, A 1968 Patrons, patronage, and political parties. Compurorive Studies in Society and Histor?: 10(4): 376-400. Wells, P. S. 1997 Identities, material culture, and change: "Celts" and "Germans" in Late Europe. Journal of European Archaeology . West, L. C. 1935 : the objects of trade. Oxford: Basi1 Blackwell. Willey, G. R 1953 Prehistoric Settlement Patterns in the Viru Valley, Peru. Bureau ofAwzerican Ethnoiogy monograph 155. Wolf, E. R. 1966 Kinship, friendship and patronclient relations in complex societies. Social arithropology of complex societies, pp. 1-22. Edited by M. Banton. New York: Praeger.